Bearing fault diagnosis method and storage medium

Abstract

The invention provides a bearing fault diagnosis method and a storage medium. The method comprises the following steps: acquiring a vibration signal of a target bearing, and performing modal decomposition on the vibration signal to obtain a preset number of intrinsic modal components; extracting features of each intrinsic mode component to form a frequency band feature vector; according to each intrinsic mode component, determining a frequency band attention weight vector; modulating the frequency band feature vector by adopting the frequency band attention weight vector to obtain a frequency band fusion feature; and obtaining a fault identification result according to the frequency band fusion features. The frequency band attention weight vector is adaptively adjusted based on the vibration signal, so that the frequency band fusion feature can be adaptively modulated, the frequency band fusion feature is more sensitive to the fault and is not sensitive to the working condition, and the accuracy of fault diagnosis is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical fault diagnosis, and in particular to a bearing fault diagnosis method and a storage medium. Background Art

[0002] Bearings are key rotating components in industrial machinery, and their health plays a vital role in the continuity and safety of the entire mechanical system. Real-time monitoring and fault diagnosis of bearing operating conditions are particularly crucial in risk-sensitive industries such as aerospace, high-speed rail, wind power generation, and intelligent manufacturing. As industrial machinery becomes increasingly information-based and intelligent, higher requirements are placed on bearing fault diagnosis methods in terms of accuracy, intelligence, generalizability, engineering practicality, and reliability.

[0003] In the existing technology, deep learning methods are widely used in the field of fault diagnosis due to their powerful nonlinear deep feature learning capabilities. However, the characteristic distribution of bearing vibration signals collected under variable operating conditions is inconsistent, and the model training set is difficult to cover all operating conditions. When the test set conditions are not included in the training set conditions, the training set and the test set do not meet the independent and identically distributed conditions, resulting in poor generalization ability of the model and low fault diagnosis accuracy, which cannot guarantee the accurate identification of bearing faults. Summary of the Invention

[0004] Embodiments of the present invention provide a bearing fault diagnosis method and storage medium to solve the problem of low fault identification accuracy of existing deep learning algorithms for bearings with variable working conditions.

[0005] In a first aspect, an embodiment of the present invention provides a bearing fault diagnosis method, comprising: Obtain the vibration signal of the target bearing and perform modal decomposition on the vibration signal to obtain a preset number of eigenmodal components; Extract the characteristics of each eigenmode component to form a frequency band feature vector; Determine the frequency band attention weight vector based on each eigenmode component; The frequency band attention weight vector is used to modulate the frequency band feature vector to obtain the frequency band fusion feature; According to the frequency band fusion characteristics, the fault identification result is obtained.

[0006] In a second aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the bearing fault diagnosis method provided in the first aspect or any possible implementation of the first aspect.

[0007] An embodiment of the present invention provides a bearing fault diagnosis method and storage medium. The above-mentioned bearing fault diagnosis method includes: obtaining the vibration signal of the target bearing, and performing modal decomposition on the vibration signal to obtain a preset number of intrinsic modal components; extracting the characteristics of each intrinsic modal component to form a frequency band feature vector; determining the frequency band attention weight vector according to each intrinsic modal component; using the frequency band attention weight vector to modulate the frequency band feature vector to obtain a frequency band fusion feature; obtaining a fault identification result according to the frequency band fusion feature. The present invention adaptively adjusts the frequency band attention weight vector based on the vibration signal, that is, adaptively adjusts it based on the specific working conditions, so that the frequency band fusion feature is insensitive to the working conditions, but more sensitive to faults, has a strong generalization ability, and effectively improves the accuracy of bearing fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0009] Figure 1 This is a flow chart of an implementation method of a bearing fault diagnosis method provided by an embodiment of the present invention; Figure 2 1 is a flow chart of an adaptive dilation rate multi-scale dilated convolution module provided by an embodiment of the present invention; Figure 3 1 is a flow chart of frequency band fusion feature extraction and fault diagnosis according to an embodiment of the present invention; Figure 4 1 is a schematic structural diagram of a bearing fault diagnosis device provided by an embodiment of the present invention; Figure 5 It is a schematic diagram of a terminal device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0010] In the following description, specific details such as particular system structures and techniques are provided for purposes of illustration, not limitation, to facilitate a thorough understanding of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

[0011] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below with reference to the accompanying drawings.

[0012] See also Figure 1 , which shows a flow chart of an implementation of a bearing fault diagnosis method provided by an embodiment of the present invention, and is described in detail as follows: The above-mentioned bearing fault diagnosis method includes: S101: Acquire a vibration signal of a target bearing, and perform modal decomposition on the vibration signal to obtain a preset number of eigenmodal components; In the embodiment of the present invention, a one-dimensional vibration signal of a fixed-length axle box in the vertical direction may be collected for analysis.

[0013] When a bearing fails, a vibration impact component will be generated in its vibration signal. When the speed and load change, the time interval of the vibration impact, the frequency and amplitude of the impact component will also change accordingly. The characteristics in the time domain are that the fault characteristics appear on different time scales at different speeds, and the characteristics in the frequency domain are that the fault characteristics appear in different frequency bands at different speeds. Since the working conditions change randomly, it is impossible to determine which time scale or frequency band feature is more sensitive to bearing failure under the current working conditions. Therefore, this application first uses the empirical mode decomposition algorithm (EMD) to perform modal decomposition on the vibration signal, and adaptively decomposes the vibration signal into a series of intrinsic mode components. .

[0014] Because the first few eigenmode components often represent the main characteristics of the signal, the dominant modes (the first few eigenmode components) carry most of the signal's energy and key information, reflecting the main dynamic characteristics of the bearing. Subsequent eigenmode components often correspond to subtle, low-frequency vibrations in the system or minor details in the signal, and may even contain noise. These low-frequency, minor modes have little impact on the overall behavior of the bearing. Retaining too many unimportant modal components may introduce unnecessary complexity, obscure key characteristics, and interfere with understanding the signal's essence.

[0015] Therefore, in the embodiment of the present invention, only a preset number of eigenmode components are retained, and the signal components of different frequency bands contained in the vibration signal are adaptively screened out from high to low step by step to form (For example, ) eigenmode components of different frequency bands, one eigenmode component corresponds to a component signal of a frequency band, so that the signal can be decomposed into Frequency bands can effectively remove noise and secondary information, simplify the analysis process, and more clearly grasp the core characteristics of the bearing. While meeting the requirements of feature analysis, the computational complexity is reduced and storage space is saved.

[0016] Since the frequency components contained in each eigenmode component change with the vibration signal, it is adaptive.

[0017] In a possible implementation, the preset number may be 5.

[0018] Based on the application scenario of the present application, the preset number may be 5. Specifically, the number of eigenmode components may be set according to actual application requirements and is not specifically limited here.

[0019] EMD is an analysis method based on the time scale of the data itself. It can decompose a complex signal into several intrinsic mode functions (IMFs) and a residual component. As an adaptive signal decomposition method, empirical mode decomposition can effectively decompose non-stationary signals, extracting useful information implicit in the signal while also performing noise reduction.

[0020] The specific steps of the EMD algorithm are: Assume the input signal is , using the cubic spline function to find the input signal The upper envelope of the signal is obtained by fitting the maximum and minimum values ​​of and lower envelope .

[0021] Average the upper and lower envelopes to get the envelope mean

[0022] For input signal and envelope mean Subtract and the remaining signal is

[0023] For the remaining signals Repeat the above process until the modal component condition SD is less than the threshold value, and the first IMF component is obtained. .

[0024] Among them, the SD condition is:

[0025] For input signal and Take the difference and get the first-order residual ,Will Replace input signal Repeat the above steps After the second you can get order eigenmode components and Residual of order .

[0026]

[0027] Therefore, the input signal The expression of EMD decomposition is: .

[0028] Input signal The length is After empirical mode decomposition, multiple IMF components are obtained, and the length of each IMF component is .

[0029] In accordance with the embodiment of the present invention, another preset number is 5, that is, , get the first five eigenmode components .

[0030] S102: extracting the features of each eigenmode component to form a frequency band feature vector; The present application extracts the features of each eigenmode component, that is, the features of each frequency band, to form a frequency band feature vector.

[0031] In a possible implementation, S102 may include: S1021: Determine a target expansion rate based on the vibration signal; S1022: Setting the dilation rate of the multi-scale dilated convolution module to a target dilation rate to obtain an adaptive dilation rate multi-scale dilated convolution module; S1023: For any eigenmode component, input the eigenmode component into an adaptive dilation rate multi-scale dilated convolution module to obtain the characteristics of the eigenmode component; The multi-scale dilated convolution module can capture multi-scale features, expand the receptive field, reduce parameters and computational complexity, and alleviate information loss in the pooling layer. Therefore, a multi-scale dilated convolution module is used in the embodiment of the present invention to extract features of different frequency bands.

[0032] The dilation rate in the multi-scale dilated convolution module is a key parameter, which determines the size of the convolution kernel in the dilated convolution. A small dilation rate is suitable for capturing local detail features, while a large dilation rate helps to obtain more global and large-scale features. Selecting an appropriate dilation rate has an important impact on the receptive field and feature extraction ability of the convolution operation. When the rotational speed increases, the fault feature frequency of the signal will increase, and features of a smaller time scale should be extracted. In this case, a smaller dilation rate should be used. When the rotational speed decreases, the fault feature frequency of the signal will decrease, and features of a larger time scale should be extracted. In this case, a larger dilation rate should be used. In the embodiment of the present invention, the target dilation rate is determined according to the vibration signal as the dilation rate of the multi-scale dilated convolution module, so that the dilation rate of the multi-scale dilated convolution module is adaptively adjusted according to the characteristics of the bearing vibration signal under different working conditions, which can optimize the network's ability to extract features of bearing vibration signals under different working conditions.

[0033] In a possible implementation, S1021 may include: 1. Determine the center of gravity frequency based on the vibration signal; 2. Input the centroid frequency into a fully connected layer and round the output to get the target expansion rate.

[0034] In one possible implementation, the center of gravity frequency The calculation formula may include:

[0035] in, The vibration signal spectral lines, , is the number of spectral lines, For the The frequency value of the spectral line.

[0036] Specifically, a certain IMF component , an adaptive expansion rate multi-scale hole convolution module is used to extract the characteristics of the signal in this frequency band , the target expansion rate is , the specific steps are as follows: refer to Figure 2 , No. Eigenmode components After a parameter (in, is the number of convolution kernels, is the width of the convolution kernel) of the Conv1D layer, the output size is Features Among them, in order to achieve the effect of denoising and reducing the feature length, the width of the convolution kernel is A larger value can be used.

[0037] The features Input into three parallel Conv1D layers, the convolution kernel widths of these three parallel Conv1D layers are different, respectively 、 and , convolution kernels of different widths are used to extract features of different scales, Changes with the vibration signal, that is, when the vibration signal changes, 、 and The value of will also change, and features of different time scales can be extracted for different vibration signals.

[0038] For example, under the same bearing fault condition, when the speed increases, the fault characteristic frequency of the vibration signal will increase, and the center of gravity frequency of the vibration signal will usually increase as well. The value changes, which results in the width of the convolution kernel of the three parallel Conv1D layers 、 and The adaptive expansion rate multi-scale hole convolution module can adaptively extract multi-scale features within the appropriate scale range for vibration signals of different speeds. The number of convolution kernel channels of the three parallel Conv1D layers is , the convolution step size of the Conv1D layer is 1. Each Conv1D layer in the adaptive expansion rate multi-scale hole convolution module can extract features of one scale, and the features of each scale are represented as 、 and In order to make the features of each scale have the same length, zero padding is used during convolution. 、 and The dimensions are .

[0039] Use a stitching layer to 、 and Concatenate along the channel dimension into a multi-scale feature , expressed as:

[0040] in, The size is , The concatenation layer can aggregate features of all scales to form a multi-scale feature set.

[0041] Will Input into a Time-Distributed Conv1D (TDConv1D) layer and output a size of The eigenvector of , in the TDConv1D layer, Conv1D is applied to of The convolution kernel width of the TDConv1D layer is , the number of channels is 1. Finally, Middle c The features of the channels are compressed into feature vectors An element in . Input into 2 fully connected layers and output multi-scale channel attention weight vector , multi-scale channel attention weight vector It is a size of vector, The c Elements represent Middle c The sensitivity of the characteristics of each channel to bearing faults.

[0042] Utilize multi-scale channel attention weight vector Multi-scale features Perform optimized modulation, and the calculation formula is:

[0043] in, Represents an element-wise product operation.

[0044] In order to optimize the multi-scale features after modulation Further extract multi-scale fusion features from Input into a Conv1D layer for feature extraction and output the features within the frequency band The purpose of this layer is to learn the dependencies between features of different scales and channels, and to mine the features of the signal as a whole. The parameters of the Conv1D layer are expressed as , the convolution step size is . And a maximum pooling layer with a pooling size of 2 is used to reduce the size of the feature. Finally, the layer outputs a size of Features , that is, The characteristics of the eigenmode components; among them, .

[0045] S1024: Forming frequency band feature vectors from the features of each eigenmode component.

[0046] For example, reference Figure 3 , corresponding to the first five eigenmode components , the characteristics of each eigenmode component form the frequency band eigenvector .

[0047] S103: Determine a frequency band attention weight vector based on each eigenmode component; The frequency band attention weight vector reflects the relative importance of different frequency band signals in the entire signal. Important frequency bands are assigned higher weights, causing the model to pay more attention to their characteristics, thereby better capturing and utilizing information valuable to the task. Relatively unimportant frequency bands are assigned lower weights, reducing the model's attention to them and thus reducing the impact of noise or irrelevant information on model performance.

[0048] Kurtosis and energy are two important concepts in signal analysis, describing signal characteristics from different perspectives. Kurtosis is particularly sensitive to pulse signals caused by faults, especially in the early stages of a fault, where the kurtosis value changes significantly. Signal energy is a measure of signal strength. When a bearing fault occurs, the energy in the frequency band containing the fault's characteristic frequency increases significantly. The embodiment of the present invention comprehensively considers multiple features such as kurtosis and energy, which can provide a more comprehensive understanding of the characteristics and status of the signal. It is more reasonable to further calculate the frequency band attention weight vector based on the kurtosis and energy features.

[0049] In a possible implementation, S103 may include: S1031: For any eigenmode component, determine the kurtosis index value of the eigenmode component and normalize it to obtain the normalized kurtosis of the eigenmode component; determine the energy index value of the eigenmode component and normalize it to obtain the normalized energy of the eigenmode component; add the normalized kurtosis and the normalized energy of the eigenmode component to obtain the attention weight of the eigenmode component; S1032: Obtain a frequency band attention weight vector based on the attention weights of each eigenmodal component.

[0050] In a possible implementation, the calculation formula for normalized kurtosis may include:

[0051]

[0052] in, For the The kurtosis index value of the eigenmode component, For the Normalized kurtosis of the eigenmode components; For the The root mean square value of the eigenmode components; For the The first eigenmode component data points; , is the preset quantity; , For analysis points Normalized energy The calculation formula may include:

[0053]

[0054] in, For the The energy index value of the eigenmode component, For the The normalized energy of the eigenmode components. Exemplary, reference Figure 3 , 5 eigenmode components Calculate the 5 kurtosis index values , normalized to get 5 normalized kurtosis ; 5 eigenmode components Calculate 5 energy index values , normalized to get 5 normalized energies , the corresponding addition results in 5 attention weights .

[0055] Among them, normalization is used to ensure that the sum of the kurtosis index of the five eigenmode components is 1, and to ensure that the sum of the energy index of the five eigenmode components is 1.

[0056] In a possible implementation, S1032 may include: 1. The attention weights of each eigenmodal component are formed into an initial attention weight vector; 2. Input the initial attention weight vector into two fully connected layers to obtain the frequency band attention weight vector.

[0057] In the embodiment of the present invention, reference Figure 3 , the attention weights of each eigenmodal component form the initial attention weight vector ,Will Input two fully connected layers to get the frequency band attention weight vector .

[0058] The calculation formula is:

[0059] in, For the Sigmoid activation function, it can be Implement a gating mechanism; and are the parameters of the two fully connected layers respectively; and are bias terms respectively; The The elements represent the frequency band feature vector Middle Characteristics of the frequency band Sensitivity to bearing failure.

[0060] refer to Figure 3 , frequency band attention weight vector consists of model parameters (e.g., parameters of two fully connected layers) and Joint decision, and The bearing fault identification method provided by the embodiment of the present invention is formed into a model. After the model is trained, the parameters of the model are fixed, so the frequency band attention weight vector in the trained model is Just follow , that is, the frequency band attention weight vector It only changes with the change of the model vibration signal. Therefore, the frequency band attention weight vector Better adaptability than fixed weights.

[0061] S104: Using the frequency band attention weight vector to modulate the frequency band feature vector to obtain the frequency band fusion feature; In a possible implementation, S104 may include: S1041: Multiply each element in the frequency band attention weight vector by each element in the frequency band feature vector, and concatenate the multiplied elements to obtain the frequency band fusion feature; Among them, the number of elements in the frequency band attention weight vector is the same as the number of elements in the frequency band feature vector.

[0062] refer to Figure 3 , using the frequency band attention weight vector Frequency band feature vector Modulation is performed, and the calculation formula is as follows:

[0063] in, It is the frequency band fusion feature.

[0064] Frequency band feature vector Adaptively modulated to obtain the frequency band fusion characteristics after modulation optimization Band attention weight vector Each element in is a value between 0 and 1, so the frequency band attention weight vector It can be regarded as a control gate that controls the amount of information transmitted from each frequency band feature to the subsequent layer, so as to enhance the frequency band features that are sensitive to faults and suppress the frequency band features that are sensitive to working conditions and useless features. More sensitive to faults, but less sensitive to changes in operating conditions.

[0065] S105: Obtain a fault identification result based on the frequency band fusion characteristics.

[0066] refer to Figure 3 , the frequency band fusion features can be The input is fed into two stacked convolutional modules for further feature extraction. The goal is to learn the dependencies between features in different frequency bands and to explore the overall characteristics of the signal. Each convolutional module contains one convolutional layer, one batch normalization layer, one activation layer, and one pooling layer. The extracted features are flattened by the flattening layer and then output as classification results using two fully connected layers and a softmax layer. y The calculation formula of the softmax output layer is:

[0067] in, Indicates the number of fault categories; Represents the parameter set of the softmax classification layer; Represents the input of the softened maximum layer; Indicates the classification result of the softened maximum layer output.

[0068] In the embodiment of the present invention, the value of the adaptive expansion rate is established to establish an adaptive expansion rate multi-scale void convolution module to extract the features of each frequency band, and at the same time, the attention weight of each frequency band is adaptively adjusted to realize the fault feature extraction under unchanged working conditions, thereby improving the generalization ability of the working condition domain of fault identification, improving the accuracy of bearing fault diagnosis under variable working conditions, and at the same time improving the interpretability and credibility of the fault diagnosis.

[0069] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0070] The following are device embodiments of the present invention. For details not fully described therein, reference may be made to the corresponding method embodiments described above.

[0071] Figure 4 The following is a schematic diagram showing the structure of a bearing fault diagnosis device provided by an embodiment of the present invention. For ease of explanation, only the parts related to the embodiment of the present invention are shown, which are described in detail as follows: like Figure 4As shown, the bearing fault diagnosis device includes: The modal decomposition module 21 is used to obtain the vibration signal of the target bearing and perform modal decomposition on the vibration signal to obtain a preset number of eigenmodal components; The initial feature extraction module 22 is used to extract the features of each intrinsic mode component to form a frequency band feature vector; The weight adaptive adjustment module 23 is used to determine the frequency band attention weight vector according to each eigenmode component; A feature modulation module 24 is used to modulate the frequency band feature vector using the frequency band attention weight vector to obtain a frequency band fusion feature; The fault identification module 25 is configured to obtain a fault identification result based on the frequency band fusion feature.

[0072] In a possible implementation, the initial feature extraction module 22 may include: an expansion rate determination unit, configured to determine a target expansion rate based on the vibration signal; A dilated convolution model building unit is used to set the dilation rate of the multi-scale dilated convolution module to the target dilation rate, thereby obtaining a multi-scale dilated convolution module with adaptive dilation rate; A feature component extraction unit is used to input any eigenmode component into an adaptive expansion rate multi-scale dilated convolution module to obtain the feature of the eigenmode component; The feature vector forming unit is used to form a frequency band feature vector from the features of each eigenmode component.

[0073] In a possible implementation, the expansion rate determination unit may be specifically configured to: 1. Determine the center of gravity frequency based on the vibration signal; 2. Input the centroid frequency into a fully connected layer and round the output to get the target expansion rate.

[0074] In one possible implementation, the center of gravity frequency The calculation formula may include:

[0075] in, The vibration signal spectral lines, , is the number of spectral lines, For the The frequency value of the spectral line.

[0076] In a possible implementation, the weight adaptive adjustment module 23 may include: a weight component calculation unit, configured to determine, for any eigenmode component, a kurtosis index value of the eigenmode component and normalize the value to obtain a normalized kurtosis of the eigenmode component; determine an energy index value of the eigenmode component and normalize the value to obtain a normalized energy of the eigenmode component; and add the normalized kurtosis and the normalized energy of the eigenmode component to obtain an attention weight of the eigenmode component; The weight vector forming unit is used to obtain a frequency band attention weight vector according to the attention weights of each eigenmode component.

[0077] In a possible implementation, the calculation formula for normalized kurtosis may include:

[0078]

[0079] in, For the The kurtosis index value of the eigenmode component, For the Normalized kurtosis of the eigenmode components; For the The root mean square value of the eigenmode components; For the The first eigenmode component data points; , is the preset quantity; , For analysis points Normalized energy The calculation formula may include:

[0080]

[0081] in, For the The energy index value of the eigenmode component, For the The normalized energy of the eigenmode components.

[0082] In a possible implementation, the weight vector forming unit may be specifically configured to: 1. The attention weights of each eigenmodal component are formed into an initial attention weight vector; 2. Input the initial attention weight vector into two fully connected layers to obtain the frequency band attention weight vector.

[0083] In a possible implementation, the feature modulation module 24 may include: The fusion and splicing unit is used to multiply each element in the frequency band attention weight vector with each element in the frequency band feature vector, and splice the multiplied elements to obtain the frequency band fusion feature; Among them, the number of elements in the frequency band attention weight vector is the same as the number of elements in the frequency band feature vector.

[0084] In a possible implementation, the preset number is 5.

[0085] Figure 5 Schematic diagram of the terminal device 3 provided by the embodiment of the present invention. Figure 5 As shown, the terminal device 3 of this embodiment includes: a processor 30 and a memory 31. The memory 31 is used to store a computer program 32, and the processor 30 is used to call and run the computer program 32 stored in the memory 31 to perform the steps in the above-mentioned various bearing fault diagnosis method embodiments, such as Figure 1 Alternatively, the processor 30 is used to call and run the computer program 32 stored in the memory 31 to implement the functions of each module / unit in the above-mentioned device embodiments, such as Figure 4 The functions of modules 21 to 25 are shown.

[0086] Exemplarily, the computer program 32 may be divided into one or more modules / units, one or more modules / units being stored in the memory 31 and executed by the processor 30 to implement the present invention. One or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program 32 in the terminal device 3. For example, the computer program 32 may be divided into Figure 4 Modules / units 21 to 25 are shown.

[0087] The terminal device 3 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal device 3 may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that Figure 5 It is only an example of terminal device 3 and does not constitute a limitation on terminal device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal may also include input and output devices, network access devices, buses, etc.

[0088] The processor 30 may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0089] The memory 31 can be an internal storage unit of the terminal device 3, such as the terminal device's hard drive or memory. Alternatively, the memory 31 can be an external storage device of the terminal device 3, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the memory 31 can include both the terminal device's internal storage unit and an external storage device. The memory 31 is used to store computer programs and other programs and data required by the terminal. The memory 31 can also be used to temporarily store data that has been output or is about to be output.

[0090] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0091] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0092] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0093] In the embodiments provided herein, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For example, the division of modules or units is merely a logical functional division. In actual implementation, other division methods may be used, such as multiple units or components being combined or integrated into another system, or some features being ignored or not implemented. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interface, or the indirect coupling or communication connection of devices or units may be electrical, mechanical, or other forms.

[0094] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0095] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0096] If the integrated module / unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the process steps in the above-mentioned method embodiments by using a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. Computer-readable media can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, mobile hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunications signals, and software distribution media.

[0097] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A bearing fault diagnosis method, characterized in that: include: Acquire a vibration signal of a target bearing, and perform modal decomposition on the vibration signal to obtain a preset number of eigenmodal components; Extract the characteristics of each eigenmode component to form a frequency band feature vector; According to each eigenmode component, determine the frequency band attention weight vector; The frequency band feature vector is modulated by using the frequency band attention weight vector to obtain a frequency band fusion feature; A fault identification result is obtained according to the frequency band fusion characteristics.

2. The bearing fault diagnosis method according to claim 1, characterized in that: The step of extracting the features of each eigenmode component to form a frequency band feature vector includes: determining a target expansion rate according to the vibration signal; The expansion rate of the multi-scale dilated convolution module is set to the target expansion rate to obtain an adaptive expansion rate multi-scale dilated convolution module; For any eigenmode component, input the eigenmode component into the adaptive dilation rate multi-scale dilated convolution module to obtain the characteristics of the eigenmode component; The characteristics of each eigenmode component are used to form the frequency band characteristic vector.

3. The bearing fault diagnosis method according to claim 2, characterized in that: Determining a target expansion rate according to the vibration signal includes: determining a center of gravity frequency based on the vibration signal; The centroid frequency is input into a fully connected layer, and the output result is rounded to obtain the target expansion rate.

4. The bearing fault diagnosis method according to claim 3, characterized in that: The center of gravity frequency The calculation formula include: in, is the vibration signal Spectrum lines, , is the number of spectral lines, For the The frequency value of the spectral line.

5. The bearing fault diagnosis method according to any one of claims 1 to 4, characterized in that: The step of determining the frequency band attention weight vector according to each eigenmode component includes: For any eigenmode component, determine the kurtosis index value of the eigenmode component and normalize it to obtain the normalized kurtosis of the eigenmode component; determine the energy index value of the eigenmode component and normalize it to obtain the normalized energy of the eigenmode component; add the normalized kurtosis and normalized energy of the eigenmode component to obtain the attention weight of the eigenmode component; According to the attention weights of each eigenmodal component, the frequency band attention weight vector is obtained.

6. The bearing fault diagnosis method according to claim 5, characterized in that: The calculation formula of the normalized kurtosis includes: in, For the The kurtosis index value of the eigenmode component, For the The normalized kurtosis of the eigenmode components; For the The RMS value of the eigenmode components; For the The first eigenmode component data points; , is the preset number; , To analyze the points; The normalized energy The calculation formula includes: in, For the The energy index value of the eigenmode component, For the The normalized energy of the eigenmode components.

7. The bearing fault diagnosis method according to claim 5, characterized in that: The step of obtaining the frequency band attention weight vector according to the attention weights of each eigenmode component includes: The attention weights of each intrinsic modal component are formed into an initial attention weight vector; The initial attention weight vector is input into two fully connected layers to obtain the frequency band attention weight vector.

8. The bearing fault diagnosis method according to any one of claims 1 to 4, characterized in that: The step of using the frequency band attention weight vector to modulate the frequency band feature vector to obtain a frequency band fusion feature includes: Multiplying each element in the frequency band attention weight vector with each element in the frequency band feature vector respectively, and concatenating the multiplied elements to obtain the frequency band fusion feature; Among them, the number of elements in the frequency band attention weight vector is the same as the number of elements in the frequency band feature vector.

9. The bearing fault diagnosis method according to any one of claims 1 to 4, characterized in that: The preset number is 5.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the bearing fault diagnosis method as described in any one of claims 1 to 9 are implemented.

Images